Proc. Nati. Acad. Sci. USA
Vol. 88, pp. 380-384, January 1991 Neurobiology
Action potential broadening and frequency-dependent facilitation of calcium signals in pituitary nerve terminals (neurohypophysis/stlmulus-ecretion coupling/neuropeptide/hormone secretion/synaptic plasticity)
MEYER B. JACKSON*, ARTHUR KONNERTHtt, AND GEORGE J. AUGUSTINE§ tMax-Planck-Institut fur Biophysikalische Chemie, D-3400 Gottingen, Federal Republic of Germany; *Department of Physiology, University of Wisconsin, Madison, WI 53706; and §Neurobiology Section, Department of Biological Sciences, University of Southern California, Los Angeles, CA 90089-2520 Communicated by Erwin Neher, October 11, 1990 (received for review July 19, 1990)
ABSTRACT Hormone release from nerve terminals in the neurohypophysis Is a sensitive function of action potential frequency. We have investigated the cellular mechanisms responsible for this frequency-dependent facilitation by combining patch clamp and fluorimetric Ca2+ measurements in single neurosecretory terminals in thin slices of the rat posterior pituitary. In these terminals both action potential-induced changes in the intracellular Ca2+ concentration ([Ca2+]J) and action potential duration were enhanced by high-frequency stimuli, all with a frequency dependence similar to that of hormone release. Furthermore, brief voltage clamp pulses inactivated a K+ current with a very similar frequency dependence. These results support a model for frequency-dependent facilitation in which the inactivation of a K+ current broadens action potentials, leading to an enhancement of [Ca2+]J signals. Further experiments tested for a causal relationship between action potential broadening and facilitation of [Ca2+J1 changes. First, increasing the duration of depolarization, either by broadening action potentials with the K+-channel blocker tetraethylammonium or by applying longer depolarizing voltage damp steps, increased [Ca2+J1 changes. Second, eliminating frequency-dependent changes in duration, by voltage clamping the terminal with constant duration pulses, substantially reduced the frequency-dependent enhancement of [Ca2+], changes. These results indicate that action potential broadening contributes to frequency-dependent facilitation of [Ca2+]J changes. However, the small residual frequency dependence of [Ca2+J1 changes seen with constant duration stimulation suggests that a second process, distinct from action potential broadening, also contributes to facilitation. These two frequency-dependent mechanisms may also contribute to activitydependent plasticity in synaptic terminals.
terminals release the neuropeptides vasopressin and oxytocin and have received much attention as a model system for the study of secretion (6-12). Peptide secretion from these terminals also exhibits a striking form of use-dependent plasticity: secretion is a sensitive function of action potential frequency (11, 13-17). This frequency dependence is essential to the input-output properties of the hypothalamichypophyseal axis (18) and has been proposed to be due to frequency-dependent modulation of action potential duration (10, 11, 13, 14, 19-23). In the present study, we use patch clamp techniques and fluorimetric calcium indicators to test this hypothesis in individual nerve terminals in pituitary slices. A preliminary account of this work has appeared (24). METHODS Neurohypophysial Slices. The neurointermediate lobe of the pituitary was removed from male rats with ages ranging from 2 to 4 months and placed in carbogen (95% 02/5% C02)saturated rat Ringer's solution (125 mM NaCl, 2.5 mM KCl, 1.25 mM NaH2PO4, 26 mM NaHCO3, 2 mM CaCl2, 1 mM MgCl2, and 20 mM glucose) at 0WC for at least 1 min. Slices 70-80 j&m thick were then cut with an FTB vibratome, with continued bathing in 0WC rat Ringer's solution. Slices were either kept in a 34WC bath or transferred to a recording chamber at room temperature (20-220C) in rat Ringer's solution identical to that described above, except that the KCl concentration was 4 mM. Slices were suitable for recording immediately and were viable for up to 4 hr. Patch clamp recordings from these slices were made at room temperature (20-220C) following published methods (25), except that no cleaning of tissue or enzyme treatment was necessary. Patch electrodes with resistances between 2.5 and 6 Mfl were fabricated from borosilicate glass (i.d. = 1.4 mm; o.d. = 2.0 mm). Tight-seal intracellular recordings (26) were achieved with series resistances ranging from 4.5 to 15 MW. For experiments in which [Ca2+], was not simultaneously measured, patch pipettes were filled with a solution consisting of 140 mM KCl, 10 mM EGTA, 4 mM ATP, 4 mM MgCI2, and 10 mM Hepes at pH 7.3. Signals were recorded with an EPC-7 patch clamp amplifier and digitized and stored on an Atari computer. A computer program written by Michael Pusch (University of Genoa) was used to acquire data. Intracellular Ca2l Measurements. For intracellular Ca2` concentration ([Ca2+]o) measurements from single terminals, the fluorescent calcium indicator fura-2 (27) (100 or 300 ,uM) was added to a patch pipette filling solution with the composition 135 mM KCl, 5 mM NaCl, 0.2 mM EGTA, 4 mM ATP, 4 mM MgCl2, and 10 mM Hepes at pH 7.3. The ratiometric technique used to measure [Ca2+], was identical to that described by Neher (28). In brief, terminals were
Little is known about the mechanisms that regulate secretion from nerve terminals, despite the importance of these mechanisms to many physiological processes. For example, a presynaptic site has been proposed as an important locus for change in several forms of use-dependent synaptic plasticity, including facilitation and posttetanic potentiation at neuromuscular synapses (1) and long-term potentiation in the mammalian hippocampus (2, 3). However, in these and most other systems one can do little more than infer that some presynaptic change occurs, because technical difficulties preclude a direct physiological characterization of the nerve terminals. Although a direct physiological characterization is possible in some invertebrate systems (4, 5), until recently there was no comparable vertebrate system for the study of presynaptic physiology. In an effort to bridge this technical gap, some investigators have made electrical and optical recordings from the nerve terminals of the vertebrate neurohypophysis (6-12). These The publication costs of this article were defrayed in part by page charge payment. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. §1734 solely to indicate this fact.
Abbreviation: TEA, tetraethylammonium. tTo whom reprint requests should be addressed. 380
Neurobiology: Jackson et al. alternately excited with UV light, filtered at 360 or 390 nm with a wheel rotating at 5 Hz, and the emitted fluorescence was measured with a Hamamatsu photomultiplier tube (model R928). [Ca2+], was then computed from the ratio of emission at the two wavelengths. Such measurements of [Ca2+]i could be made from single terminals for up to 30 min.
RESULTS Nerve terminals were visible in thin slices of the neurohypophysis when viewed under Nomarski optics (Fig. 1). These terminals are round in appearance, with diameters ranging from I to 20 gm. The identity of these structures as terminals has been established (S. A. DeRiemer, R. Schneggenburger, M.B.J., and A.K., unpublished results) by the following criteria: (i) absence of nuclei, as judged by the lack of staining with a nuclear dye; (ii) orthograde transport of dye from hypothalamic neurons; and (iii) axonal stimulation generates action potentials. Furthermore, as shown below, electrical stimulation leads to increases in [Ca2+li in these structures. The terminals observed on the upper surface of a slice were often exposed and accessible to patch pipettes. An example of such a terminal, filled with the fluorescent dye Lucifer yellow, is shown in Fig. 1. Calcium Signaling in Single Nerve Terminals. Loading of the fluorescent Ca2" indicator fura-2 (27) from the patch pipette solution was accomplished in 2 min or less. The resting [Ca2+], in these terminals was 110 ± 10 nM (mean ± SEM; n = 16), which is comparable to the resting [Ca2+]i of many cell types (29). The value of 350 nM reported for secretosomes prepared from the posterior pituitary is considerably higher (30), but this discrepancy can be explained by leakage of dye from secretosomes. Since secretion from the nerve terminals of the posterior pituitary is triggered by Ca2" entry (8, 9, 31), we were
Proc. Natl. Acad. Sci. USA 88 (1991)
381
interested in asking whether the frequency dependence of secretion was a result of the frequency dependence of [Ca2"J changes. Repetitive stimulation of current-clamped terminals with current pulses large enough to evoke action potentials caused transient increases in [Ca2+Ji (Fig. 2). These increases were rapid, with a rate of rise that depended on the duration and frequency of stimulation. The peak in [Ca2+]i was followed by an exponential decay back to the resting level. The half-time for decay was in the range of 10 to 20 s. Ca2' buffers such as fura-2 are known to slow [Ca2+]j recovery (ref. 32; G.J.A. and E. Neher, unpublished results), so that the fura-2 may have slowed the decay of [Ca2+], signals. However, lowering the fura-2 concentration from 300 to 100 /iM had little effect on the time course of recovery, which suggests that such effects were relatively minor in our experiments. Stimulation with action potentials at high frequencies increased [Ca2+]J much more than stimulation with action potentials at low frequencies (Fig. 2). At 1 Hz the change in [Ca2+], induced by 100 action potentials was barely detectable, but following 100 action potentials at frequencies of 10 Hz or higher, [Ca2+]i increased to 300 nM or more (Fig. 2A). High-frequency stimulation also enhanced the rate of increase of [Ca2+]i during the stimulus train. The frequency dependence of action potential-evoked increases in [Ca2+]J is shown in Fig. 2B. This frequency dependence resembles that reported for vasopressin release from isolated neurointermediate lobes (13, 14). Both signals are maximal at =10 Hz, although the frequency of the half-maximal response is roughly 3 Hz higher for vasopressin release than for [Ca2+], signals (Fig. 2B). This difference could be due to a nonlinear relationship between [Ca2+]i and the rate of secretion (33-35) or to minor differences in experimental conditions (vasopressin release was measured in a similar saline but with a bovine serum albumin concentration of 1.5 mg/ml at 37°C). Nonetheless, the similar frequency dependences of [Ca2+]i and
FIG. 1. Nerve terminals on the surface of a thin pituitary slice. A photomicrograph with simultaneous Nomarski and fluorescence optics shows a Lucifer yellow-filled nerve terminal. The patch electrode, filled with Lucifer yellow (0.2 mg/ml) in our normal patch-electrode filling solution, is also visible. (Bar = 20 iam.)
382
Neurobiology: Jackson et al.
Proc. Natl. Acad. Sci. USA 88
A 20
Hz'- 0.2 0.1 [COal I
40 s
0
5 Hz
1 Hz
B
.-
1.,12
u
Vol. 88, pp. 380-384, January 1991 Neurobiology
Action potential broadening and frequency-dependent facilitation of calcium signals in pituitary nerve terminals (neurohypophysis/stlmulus-ecretion coupling/neuropeptide/hormone secretion/synaptic plasticity)
MEYER B. JACKSON*, ARTHUR KONNERTHtt, AND GEORGE J. AUGUSTINE§ tMax-Planck-Institut fur Biophysikalische Chemie, D-3400 Gottingen, Federal Republic of Germany; *Department of Physiology, University of Wisconsin, Madison, WI 53706; and §Neurobiology Section, Department of Biological Sciences, University of Southern California, Los Angeles, CA 90089-2520 Communicated by Erwin Neher, October 11, 1990 (received for review July 19, 1990)
ABSTRACT Hormone release from nerve terminals in the neurohypophysis Is a sensitive function of action potential frequency. We have investigated the cellular mechanisms responsible for this frequency-dependent facilitation by combining patch clamp and fluorimetric Ca2+ measurements in single neurosecretory terminals in thin slices of the rat posterior pituitary. In these terminals both action potential-induced changes in the intracellular Ca2+ concentration ([Ca2+]J) and action potential duration were enhanced by high-frequency stimuli, all with a frequency dependence similar to that of hormone release. Furthermore, brief voltage clamp pulses inactivated a K+ current with a very similar frequency dependence. These results support a model for frequency-dependent facilitation in which the inactivation of a K+ current broadens action potentials, leading to an enhancement of [Ca2+]J signals. Further experiments tested for a causal relationship between action potential broadening and facilitation of [Ca2+J1 changes. First, increasing the duration of depolarization, either by broadening action potentials with the K+-channel blocker tetraethylammonium or by applying longer depolarizing voltage damp steps, increased [Ca2+J1 changes. Second, eliminating frequency-dependent changes in duration, by voltage clamping the terminal with constant duration pulses, substantially reduced the frequency-dependent enhancement of [Ca2+], changes. These results indicate that action potential broadening contributes to frequency-dependent facilitation of [Ca2+]J changes. However, the small residual frequency dependence of [Ca2+J1 changes seen with constant duration stimulation suggests that a second process, distinct from action potential broadening, also contributes to facilitation. These two frequency-dependent mechanisms may also contribute to activitydependent plasticity in synaptic terminals.
terminals release the neuropeptides vasopressin and oxytocin and have received much attention as a model system for the study of secretion (6-12). Peptide secretion from these terminals also exhibits a striking form of use-dependent plasticity: secretion is a sensitive function of action potential frequency (11, 13-17). This frequency dependence is essential to the input-output properties of the hypothalamichypophyseal axis (18) and has been proposed to be due to frequency-dependent modulation of action potential duration (10, 11, 13, 14, 19-23). In the present study, we use patch clamp techniques and fluorimetric calcium indicators to test this hypothesis in individual nerve terminals in pituitary slices. A preliminary account of this work has appeared (24). METHODS Neurohypophysial Slices. The neurointermediate lobe of the pituitary was removed from male rats with ages ranging from 2 to 4 months and placed in carbogen (95% 02/5% C02)saturated rat Ringer's solution (125 mM NaCl, 2.5 mM KCl, 1.25 mM NaH2PO4, 26 mM NaHCO3, 2 mM CaCl2, 1 mM MgCl2, and 20 mM glucose) at 0WC for at least 1 min. Slices 70-80 j&m thick were then cut with an FTB vibratome, with continued bathing in 0WC rat Ringer's solution. Slices were either kept in a 34WC bath or transferred to a recording chamber at room temperature (20-220C) in rat Ringer's solution identical to that described above, except that the KCl concentration was 4 mM. Slices were suitable for recording immediately and were viable for up to 4 hr. Patch clamp recordings from these slices were made at room temperature (20-220C) following published methods (25), except that no cleaning of tissue or enzyme treatment was necessary. Patch electrodes with resistances between 2.5 and 6 Mfl were fabricated from borosilicate glass (i.d. = 1.4 mm; o.d. = 2.0 mm). Tight-seal intracellular recordings (26) were achieved with series resistances ranging from 4.5 to 15 MW. For experiments in which [Ca2+], was not simultaneously measured, patch pipettes were filled with a solution consisting of 140 mM KCl, 10 mM EGTA, 4 mM ATP, 4 mM MgCI2, and 10 mM Hepes at pH 7.3. Signals were recorded with an EPC-7 patch clamp amplifier and digitized and stored on an Atari computer. A computer program written by Michael Pusch (University of Genoa) was used to acquire data. Intracellular Ca2l Measurements. For intracellular Ca2` concentration ([Ca2+]o) measurements from single terminals, the fluorescent calcium indicator fura-2 (27) (100 or 300 ,uM) was added to a patch pipette filling solution with the composition 135 mM KCl, 5 mM NaCl, 0.2 mM EGTA, 4 mM ATP, 4 mM MgCl2, and 10 mM Hepes at pH 7.3. The ratiometric technique used to measure [Ca2+], was identical to that described by Neher (28). In brief, terminals were
Little is known about the mechanisms that regulate secretion from nerve terminals, despite the importance of these mechanisms to many physiological processes. For example, a presynaptic site has been proposed as an important locus for change in several forms of use-dependent synaptic plasticity, including facilitation and posttetanic potentiation at neuromuscular synapses (1) and long-term potentiation in the mammalian hippocampus (2, 3). However, in these and most other systems one can do little more than infer that some presynaptic change occurs, because technical difficulties preclude a direct physiological characterization of the nerve terminals. Although a direct physiological characterization is possible in some invertebrate systems (4, 5), until recently there was no comparable vertebrate system for the study of presynaptic physiology. In an effort to bridge this technical gap, some investigators have made electrical and optical recordings from the nerve terminals of the vertebrate neurohypophysis (6-12). These The publication costs of this article were defrayed in part by page charge payment. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. §1734 solely to indicate this fact.
Abbreviation: TEA, tetraethylammonium. tTo whom reprint requests should be addressed. 380
Neurobiology: Jackson et al. alternately excited with UV light, filtered at 360 or 390 nm with a wheel rotating at 5 Hz, and the emitted fluorescence was measured with a Hamamatsu photomultiplier tube (model R928). [Ca2+], was then computed from the ratio of emission at the two wavelengths. Such measurements of [Ca2+]i could be made from single terminals for up to 30 min.
RESULTS Nerve terminals were visible in thin slices of the neurohypophysis when viewed under Nomarski optics (Fig. 1). These terminals are round in appearance, with diameters ranging from I to 20 gm. The identity of these structures as terminals has been established (S. A. DeRiemer, R. Schneggenburger, M.B.J., and A.K., unpublished results) by the following criteria: (i) absence of nuclei, as judged by the lack of staining with a nuclear dye; (ii) orthograde transport of dye from hypothalamic neurons; and (iii) axonal stimulation generates action potentials. Furthermore, as shown below, electrical stimulation leads to increases in [Ca2+li in these structures. The terminals observed on the upper surface of a slice were often exposed and accessible to patch pipettes. An example of such a terminal, filled with the fluorescent dye Lucifer yellow, is shown in Fig. 1. Calcium Signaling in Single Nerve Terminals. Loading of the fluorescent Ca2" indicator fura-2 (27) from the patch pipette solution was accomplished in 2 min or less. The resting [Ca2+], in these terminals was 110 ± 10 nM (mean ± SEM; n = 16), which is comparable to the resting [Ca2+]i of many cell types (29). The value of 350 nM reported for secretosomes prepared from the posterior pituitary is considerably higher (30), but this discrepancy can be explained by leakage of dye from secretosomes. Since secretion from the nerve terminals of the posterior pituitary is triggered by Ca2" entry (8, 9, 31), we were
Proc. Natl. Acad. Sci. USA 88 (1991)
381
interested in asking whether the frequency dependence of secretion was a result of the frequency dependence of [Ca2"J changes. Repetitive stimulation of current-clamped terminals with current pulses large enough to evoke action potentials caused transient increases in [Ca2+Ji (Fig. 2). These increases were rapid, with a rate of rise that depended on the duration and frequency of stimulation. The peak in [Ca2+]i was followed by an exponential decay back to the resting level. The half-time for decay was in the range of 10 to 20 s. Ca2' buffers such as fura-2 are known to slow [Ca2+]j recovery (ref. 32; G.J.A. and E. Neher, unpublished results), so that the fura-2 may have slowed the decay of [Ca2+], signals. However, lowering the fura-2 concentration from 300 to 100 /iM had little effect on the time course of recovery, which suggests that such effects were relatively minor in our experiments. Stimulation with action potentials at high frequencies increased [Ca2+]J much more than stimulation with action potentials at low frequencies (Fig. 2). At 1 Hz the change in [Ca2+], induced by 100 action potentials was barely detectable, but following 100 action potentials at frequencies of 10 Hz or higher, [Ca2+]i increased to 300 nM or more (Fig. 2A). High-frequency stimulation also enhanced the rate of increase of [Ca2+]i during the stimulus train. The frequency dependence of action potential-evoked increases in [Ca2+]J is shown in Fig. 2B. This frequency dependence resembles that reported for vasopressin release from isolated neurointermediate lobes (13, 14). Both signals are maximal at =10 Hz, although the frequency of the half-maximal response is roughly 3 Hz higher for vasopressin release than for [Ca2+], signals (Fig. 2B). This difference could be due to a nonlinear relationship between [Ca2+]i and the rate of secretion (33-35) or to minor differences in experimental conditions (vasopressin release was measured in a similar saline but with a bovine serum albumin concentration of 1.5 mg/ml at 37°C). Nonetheless, the similar frequency dependences of [Ca2+]i and
FIG. 1. Nerve terminals on the surface of a thin pituitary slice. A photomicrograph with simultaneous Nomarski and fluorescence optics shows a Lucifer yellow-filled nerve terminal. The patch electrode, filled with Lucifer yellow (0.2 mg/ml) in our normal patch-electrode filling solution, is also visible. (Bar = 20 iam.)
382
Neurobiology: Jackson et al.
Proc. Natl. Acad. Sci. USA 88
A 20
Hz'- 0.2 0.1 [COal I
40 s
0
5 Hz
1 Hz
B
.-
1.,12
u
